Niki Dupuis

Documenting a failed experiment My question: could you upgrade a base model by giving it more time to think? Concretely, could you finetune a base model (pretrain only) to make effective use of filler tokens during inference. I looked around and found a few papers, but in all cases they...

There are a lot of AI Safety papers, essays, think pieces, discussions, all written in English. What if we used LLMs to translate them into every major language? Here I lay out briefly why I think this could be a good idea, and then I attempt to have Claude do...

One People When the physicists first told us about atmospheric ignition, we had to grow up. Without gods to protect us, we learned to be afraid. There could be no more war, no arms races. From now on, we had to become one people, a social organism. We held a...

This post is a distillation of a recent work in AI-assisted human coordination from Google DeepMind. The paper has received some press attention, and anecdotally, it has become the de-facto example that people bring up of AI used to improve group discussions. Since this work represents a particular perspective/bet on...

Pantheon is an experimental LLM interface exploring a different type of human-AI interaction. We created this as a part of the cyborgism project, with the abstract motivation of augmenting the human ability to think by integrating human and AI generated thoughts. How it works: 1. A human user “thinks out...

I did an exploration into how Community Notes (formerly Birdwatch) from X (formerly Twitter) works, and how its algorithm decides which notes get displayed to the wider community. In this post, I’ll share and explain what I found, as well as offer some comments. Community Notes is a fact-checking tool...

Often, when people explain the behavior of an AI system, there is one of Daniel Dennett’s stances implicit in their explanation. I think that when this happens implicitly, it can often lead to misleading claims and sloppy reasoning. For this reason, I feel it can be really valuable to be...